The present invention relates to a sampling circuit and an amplification type solid-state imaging device.
Conventionally, as an amplification type solid-state imaging device that employs a sampling circuit, there has been a known one in which each pixel element is provided with an amplification function for to read an optical reception signal by means of a scanning circuit. In particular, there have been energetically developed APS (Active Pixel Sensor) type image sensors having a CMOS type pixel element construction that is advantageous for the element to be integrated with a peripheral drive circuit and a signal processing circuit. In this APS type image sensor, it is required to form a photoelectric conversion section, an amplification section, a pixel selection section and a reset section within one pixel element, and normally three to four MOS transistors (T) are employed besides the photoelectric conversion section constructed of a photo diode (PD).
FIG. 5 shows the construction of the pixel element in the case of a PD+3T system. In FIG. 5, signal charges generated in a photoelectric conversion section 101 are converted into a voltage and applied to the gate of a transistor 102. This transistor 102 executes impedance conversion (current amplification), and a signal Vsig is read via a pixel selection switch 103 to the gate of which a voltage xcfx86x is applied. In the middle of the read period of this signal Vsig, the potential of the photoelectric conversion section 101 is reset to a power voltage VD by a reset transistor 104 to the gate of which a voltage xcfx86R is applied. Therefore, a potential (optical reception signal) corresponding to the signal charges and a reset potential (reference signal) are read as a pair into the signal Vsig.
The pixel section of the amplification type solid-state imaging device that employs the pixel element shown in FIG. 5 can be expressed by a schematic diagram as shown in FIG. 6. In FIG. 6 is shown a pixel section 131 for executing the photoelectric conversion, amplification, read and reset. The read of the pixel section 131 is controlled by a voltage xcfx86X of a signal line 106, and the reset of the pixel section 131 is controlled by a voltage xcfx86R of a signal line 107. Thus, the amplified signal Vsig is outputted from the pixel section 131 via a vertical signal line 108.
FIG. 7 is a schematic diagram of an amplification type solid-state imaging device (two-dimensional image sensor) constructed by employing the pixel section 131. In FIG. 7, a two-dimensional pixel region 140 is constituted of the pixel section 131, a first vertical scanning circuit 141 and a second vertical scanning circuit 142. The read operation of the pixel section 131 is controlled by a signal 143 from the first vertical scanning circuit 141, and the reset operation is controlled by a signal 144 from the second vertical scanning circuit 142. An output signal from the pixel section 131 is read by a vertical signal line 145 and thereafter transferred to a correlated double sampling circuit provided for each vertical signal line 145. A differential signal representing a difference between the optical reception signal obtained during read and a reference signal obtained after the reset is outputted from the correlated double sampling circuit. With regard to the differential signal, variations in the threshold value between the pixel sections 131 are canceled, so that a fixed pattern noise (hereinafter referred to as FPN) of each pixel section 131 is suppressed. It is to be noted that the correlated double sampling circuit is constituted of a clamping circuit (clamping capacitor 146 and clamping switch 147) and a sample hold circuit (sample hold switch 148 and sample hold capacitor 149).
In the aforementioned correlated double sampling circuit, the vertical signal line 145 is connected to the sample hold switch 148 via the clamping capacitor 146 and connected to the clamping potential VCP via the clamping switch 147.
FIGS. 8A through 8G show the operation timing of the aforementioned correlated double sampling circuit. In the clamping circuit (clamping capacitor 146 and clamping switch 147), the operation of clamping to the clamping potential VCP is executed by making a pulse xcfx86C1 have high level during read of the optical reception signal (S1) from the pixel section 131, and a differential signal Vsj of a difference between the optical reception signal (S1) from the pixel section 131 and the reference signal (S2) is outputted from the clamping capacitor 146. The sample hold switch 148 samples the differential signal Vsj of the difference between the optical reception signal and the reference signal. Then, the signal Vsj from the sample hold switch 148 is held by the sample hold capacitor 149 and amplified by an amplifier circuit 155. The signal amplified by the amplifier circuit 155 is outputted as an output signal Vsr to the horizontal signal line 164 via a horizontal selection switch 156 controlled by an output line 161 from the horizontal scanning circuit 160, and differential signals Vsj, Vsj+1, . . . from the sample hold circuits are successively read. It is to be noted that the reference numeral 165 denotes a second constant current load of the amplifier circuit 155. Finally, the signal Vsr becomes a signal OS by the operation of a buffer amplifier 169.
As described above, in the amplification type solid-state imaging device shown in FIG. 7, the FPN due to the variations in the threshold value between the pixel sections 131 is suppressed by the correlated double sampling circuit provided for each vertical signal line 145. However, in the amplification type solid-state imaging device, the amplifier circuits 155 of the vertical signal lines 145 are accompanied by variations in the offset level, gain and so on. These variations, which are random in the horizontal direction and common in the vertical direction of an image, cause a significant FPN of a vertical stripe pattern in a video image, significantly impairing the image quality.
Accordingly, as a method for resolving the FPN of a vertical stripe pattern described above, the present applicant has proposed an amplification type solid-state imaging device as shown in FIG. 9 (Japanese Patent Application No. 2000-49505). It is to be noted that this amplification type solid-state imaging device is described for better understanding of the present invention and is neither a known technology nor a prior art.
In this amplification type solid-state imaging device, the two-dimensional pixel region has the same construction as that of the two-dimensional pixel region 140 shown in FIG. 7, and neither of figures nor description for the two-dimensional pixel region is provided. Furthermore, in order to explain only the essential point, description is started from the output side of the clamping circuit with the detailed concrete circuit omitted.
In the case of FIG. 7, an output from the clamping circuit is only the differential signal Vsj of the difference between the optical reception signal and the reference signal with respect to the clamping potential VCP that serves as a reference. This signal is sampled once by the sample hold circuit (sample hold switch 148 and sample hold capacitor 149) and transferred to the amplifier circuit 155. On the other hand, a differential signal Vsj(a) of the difference between the optical reception signal and the reference signal is read to the output side of the clamping circuit, and thereafter a reference signal Vsj(b) is further read in the case of FIG. 9. The clamping potential, which is usually selected as this reference signal, serves as a reference for the differential signal Vsj(a) and is common to the columns.
FIGS. 10A through 10F are timing charts showing the timing of the signals of the aforementioned amplification type solid-state imaging device. As shown FIGS. 10A through 10F, in the amplification type solid-state imaging device, the sample hold circuit executes the sampling operation two times. At the first time, the differential signal Vsj(a) of the difference between the optical reception signal and the reference signal is sampled by a pulse xcfx86sj, xcfx86sj+1, . . . (FIGS. 10B and 10C). This is the timing common to the columns. This differential signal is applied to the gate of the amplifier circuit 155. The horizontal selection switches 156 are sequentially turned on by pulses xcfx86Hj, xcfx86Hj+1, . . . (FIGS. 10D and 10E) of the output line 161 from the horizontal scanning circuit 160. In the middle of each ON-state period, the sample hold circuit executes the second sampling operation as indicated by the pulse xcfx86sj, xcfx86sj+1, . . . . By this operation, the signal Vsj(a), Vsj+1(a), . . . is read in the first half of the respective ON-state period, and the reference signal Vsj(b), Vsj+1(b), . . . is read in the latter half. In FIG. 10F, the signal Vsj(a), Vsj+1(a), . . . is denoted by xe2x80x9cSaxe2x80x9d, while the signal Vsj(b), Vsj+1(b), . . . is denoted by xe2x80x9cSbxe2x80x9d. The signals Vsj(a) and Vsj(b), the signals Vsj+1(a) and Vsj+1(b) and so on include the same variations in the offset level, the gain and so on of the amplifier circuit 155. Therefore, if a difference between both the signals is taken through the subsequent signal processing, then the variations in the offset level, the gain and so on of the amplifier circuit 155 can be removed.
In the amplification type solid-state imaging device shown in FIG. 9, the horizontal selection switch 156 is off during the first sampling operation. Therefore, no read current flows through the amplifier circuit 155, and a potential VN across the amplifier circuit 155 and the horizontal selection switch 156 has a high potential in a floating state. This potential VN and the gate-side potential VM of the amplifier circuit 155 are coupled with each other by way of a gate-source capacitance CGS of the transistor of the amplifier circuit 155. Therefore, a write operation is executed in the sample hold capacitor 149 in the state shifted to the high potential during the first sampling operation.
Next, during the second sampling operation, the horizontal selection switch 156 is on. Therefore, the read current flows to the amplifier circuit 155, and the potential VN across the amplifier circuit 155 and the horizontal selection switch 156 becomes a low potential. During the second sampling operation, a write operation is executed in the sample hold capacitor 149 in the state shifted to a low potential due to the influence of the capacitance CGS between the aforementioned potential VN and the gate-side potential VM of the amplifier circuit 155.
For the above reasons, a potential shift that depends on the gate-source capacitance CGS of the transistor 155 is generated between the potentials Vsj(a), Vsj+1(a), . . . through the first sampling operation and the potentials Vsj(b), Vsj+1(b), . . . through the second sampling operation. In general, the gate-source capacitance CGS of the transistor 155 has a variation. Therefore, if a difference is taken between Vsj(a) and Vsj(b), between Vsj+1(a) and Vsj+1(a) or between a similar pair, this variation component disadvantageously remains, causing a problem that a new fixed pattern noise is generated.
Accordingly, the object of the present invention is to provide a sampling circuit that receives no influence of the variation in the input-output capacitance of the amplification means and a high-performance amplification type solid-state imaging device in which the occurrence of a fixed pattern noise is remarkably reduced.
In order to achieve the above object, there is provided a sampling circuit having
a first sample hold means having a sample hold switch whose one terminal receives an input signal and a sample hold capacitor whose one terminal is connected to the other terminal of the sample hold switch,
an amplification means connected to the other terminal of the sample hold switch, and
an output selection switch whose one terminal is connected to an output of the amplification means and the other terminal of which is connected to an output signal line, the sampling circuit comprising:
a first constant current load connected to an output side of the amplification means.
In one embodiment of the present invention, a conduction control switch is connected between the amplification means and the first constant load.
According to the sampling circuit of the above-mentioned embodiment, even if the sample hold switch is turned on during the period during which the output selection switch is off, the conduction control switch is on. Therefore, even if the output selection switch is off during the first sampling operation when the present invention is applied to the correlated double sampling circuit, a current flows from the amplification means to the first constant current load via the conduction control switch. During the second sampling operation, the output selection switch is on, and the conduction control switch is off. Therefore, a read current flows from the amplification means to the output signal line via the output selection switch. Therefore, the potential across the amplification means and the output selection switch is lowered during both the first and second sampling operations in the case of the correlated double sampling circuit. Even if the write potential during the sampling operation is influenced by the capacitance between the potential across the amplification means and the output selection switch and the gate-side potential of the amplification means, the potential across the amplification means and the output selection switch is the same during both the first and second sampling operations. If the difference between the two (between the signal sampled at the first time and the signal sampled at the second time) is taken, then the influence will be canceled. Therefore, even if the capacitance varies between the plurality of sample hold circuits, the capacitance exerts no influence. Therefore, the occurrence of a new fixed pattern noise is suppressed in the amplification type solid-state imaging device that employs this sampling circuit.
In one embodiment of the present invention, the conduction control switch is turned on during at least a period during which the sample hold switch is on and the output selection switch is off.
According to the sampling circuit of the above-mentioned embodiment, by turning on the conduction control switch during at least the period during which the sample hold switch is on and the output selection switch is off, a current flows from the amplification means to the first constant current load via the conduction control switch, and a read current flows from the amplification means to the output signal line via the output selection switch during the period during which the output selection switch is on. Therefore, during at least the period during which the sample hold switch is on and the output selection switch is off, the potential across the amplification means and the output selection switch can be lowered similarly to the period during which the output selection switch is on.
In one embodiment of the present invention, the sampling circuit further comprises a second constant current load connected to the output signal line.
According to the sampling circuit of the above-mentioned embodiment, by connecting the second constant current load to the output signal line, when the output selection switch is on and the conduction control switch is off, a read current flows from the amplification means to the second constant current load via the output selection switch and the output signal line. Therefore, during at least the period during which the sample hold switch is on and the output selection switch is off and during the period during which the output selection switch is on, the potential across the amplification means and the output selection switch can be lowered and brought into the same state.
In one embodiment of the present invention, the sampling circuit further comprises a first clamping means having a clamping capacitor whose one terminal is connected to an input signal line and the other terminal of which is connected to one terminal of the sample hold switch and a clamping switch whose one terminal is connected to the other terminal of the clamping capacitor and the other terminal of which is connected to a clamping potential.
According to the sampling circuit of the above-mentioned embodiment, the differential signal of the two signals of different input signal lines can be obtained by the correlated double sampling operation by the first clamping means that has the clamping capacitor and the clamping switch and the first sample hold means.
In one embodiment of the present invention, the sampling circuit further comprises:
a control means for controlling the sample hold switch, the output selection switch, the conduction control switch and the clamping switch,
the control means operating to
clamp in a first period either one of a pair of input signal and reference signal applied to the input signal line in a first half of the first period by means of the first clamping means, thereafter sample and hold a differential signal that represents a difference between the input signal and the reference signal and appears on an output side of the clamping capacitor by the first sample hold means in a latter half of the first period and turn on the conduction control switch during at least a period during which the first sample hold means is turned on within the first period, and operating to
hold in a second period subsequent to the first period a signal on the output side of the clamping capacitor in an early stage of the second period, thereafter turn on the output selection switch during a third period within the second period, read the differential signal to the output signal line in a first half of the third period, sample and hold a constant potential signal held in the clamping capacitor by turning on the sample hold switch about in the middle of the third period, read the constant potential signal to the output signal line in a latter half of the third period and turn off the conduction control switch in the third period.
According to the sampling circuit of the above-mentioned embodiment, the input signal and the reference signal exist as a pair on the input signal line. The control means controls the sample hold switch, the output selection switch, the conduction control switch and the clamping switch so as to clamp either one of the input signal and the reference signal on the input signal line in the first half of the first period by the first clamping means, thereafter sample and hold the differential signal that is the difference between the input signal and the reference signal and appears on the output side of the clamping capacitor by the first sample hold means in the latter half of the first period, turn on the conduction control switch during at least the period during which the first sample hold means is turned on within the first period, hold the signal on the output side of the clamping capacitor to the specified potential in the early stage of the second period subsequent to the first period, thereafter read the differential signal to the output signal line in the first half of the third period during which the output selection switch is turned on within the second period, sample and hold the constant potential signal by turning on the sample hold switch about in the middle of the third period, read the constant potential signal to the output signal line in the latter half of the third period and turn off the conduction control switch in the third period.
Therefore, even if the output selection switch is off during the first sampling operation, a current flows from the amplification means to the first constant current load via the conduction control switch. During the second sampling operation, the output selection switch is on, and the conduction control switch is off. Therefore, the read current flows from the amplification means to the output signal line via the output selection switch. Therefore, the potential across the amplification means and the output selection switch is lowered during both the first and second sampling operations. If the write potential during the sampling operation is influenced by the capacitance between the potential across the amplification means and the output selection switch and the gate-side potential of the amplification means, the potential across the amplification means and the output selection switch is in the same state during both the first and second sampling operations. If the difference between the two (between the signal sampled at the first time and the signal sampled at the second time) is taken, then the influence can be canceled.
In one embodiment of the present invention, the sampling circuit further comprises:
a second clamping means for outputting a difference signal of a difference between the differential signal and the constant potential signal by clamping the differential signal on the basis of the differential signal read to the output signal line and the constant potential signal; and
a second sample hold means for sampling and holding the difference signal from the second clamping means and outputting the difference signal that is sampled and held.
According to the sampling circuit of the above-mentioned embodiment, with the construction of the second clamping circuit and the second sample hold circuit, the difference signal of the differential signal of the pair of signals read to the horizontal signal line and the constant potential signal is obtained. In the amplification type solid-state imaging device that employs this sampling circuit, a signal that represents an image from which all the variation components are removed and is free from FPN can be obtained.
Also, there is provided an amplification type solid-state imaging device provided with an amplification type pixel element, which amplifies an optical reception signal formed by a photoelectric conversion means and a reference signal that serves as a reference for the optical reception signal, wherein
the sampling circuit is employed by connecting the output of the pixel element to the input signal line, using the optical reception signal outputted from the pixel element as an input signal and using the reference signal outputted from the pixel element as a reference signal.
According to the amplification type solid-state imaging device of the above-mentioned construction, by sampling and holding the differential signal of the difference between the input signal and the reference signal in the first period by the first sample hold means and turning on the conduction control switch during at least the period during which the first sample hold means is turned on within the first period with the optical reception signal of the pixel element made to serve as an input signal and with the reference signal made to serve as the reference signal, a current flows from the amplification means to the first constant current load via the conduction control switch even if the output selection switch is off during the first sampling operation. The signal on the output side of the clamping capacitor is held at the specified potential in the early stage of the second period subsequent to the first period, and the differential signal is thereafter read to the output signal line in the first half of the third period during which the output selection switch is turned on within the second period. The constant potential signal is sampled and held by turning on the sample hold switch about in the middle of the third period, and the constant potential signal is read to the output signal line in the latter half of the third period. During this second sampling operation, the output selection switch is on, and the conduction control switch is off. Therefore, the read current flows through the output signal line from the amplification means via the output selection switch. Therefore, during both the first and second sampling operations, the potential across the amplification means and the output selection switch is lowered. Even if the write potential during the sampling operation is influenced by the capacitance between the potential across the amplification means and the output selection switch and the gate-side potential of the amplification means, the potential across the amplification means and the output selection switch is the same during both the first and second sampling operations. If the difference between the two (between the signal sampled at the first time and the signal sampled at the second time) is taken, then the influence will be canceled. Therefore, even if the capacitance varies between the plurality of sample hold circuits, there is no influence of the capacitance. Therefore, the occurrence of a new fixed pattern noise is suppressed.